National Medal of Technology








Roger L. Easton

29 March 2005

The history of GPS is so shrouded in mistakes, exaggeration and misconceptions that I think it is worthwhile to give my personal view of its history. Since I am one of the very few who was around at the beginning, I shall write this from my point of view with appropriate references

  1. Early History

  2. The Invention







  9. Two patents are of interest.

Early History

The GPS consists of about two dozen satellites that emit signals all of which are synchronized to a common time and ultimately to Naval Observatory time. A user measures the time delay between his clock and the signals from each of the satellites. By solving the problem of finding his location and, with the data from an extra satellite he can also solve for the system time. We now have what von Moltke called “Einheitszeit” or the unity of time. It is also called a central clock, the horologe-mere, the Primare Normaluhr, and the master clock (See Galison). Whatever its name, GPS can and does provide a uniform world-wide time.

Our first use of a time-related system was in connection with a problem in time synchronization. We had a bi-static radar using continuous wave radar at two sites in Southern Texas. We had the problem of synchronizing the transmitted and received signals. We tried over-the–horizon signals and found them noisy so we carried cesium-beam clocks between the transmitter and receiver to synchronize the two equipments. This technique worked well but was time-consuming as the two stations were about 100 miles apart.

In discussing the problem at the receiving station near Harlingen, Texas in 1964 it occurred to me that one answer to the problem would be to have the time carried by a satellite clock. Accordingly, we looked into the feasibility of designing such a satellite. Later, while I was on Naval Reserve Training Duty, I had the idea of using the technique for navigation. We demonstrated the technique by having a receiving station at the Laboratory and a clock in an automobile going down route 295 before the route was opened officially. It is therefore not surprising that the first idea for a time related system was in connection with a time transfer problem. We had bi-static radar for the detection and location of satellites using continuous wave radar located at two sites in South Texas. We had the problem of synchronizing the transmitted to the received signals. We tried over-the-horizon signals and found them noisy so we carried cesium-beam clocks between transmitter and receiver to synchronize the two equipments. This technique worked well but was time consuming.

The demonstration was held for the late John Yob and for Chester Kleczek of the Naval Weapons Command (later the Naval Air Systems Command). As a result of its success we obtained a work order of $35k to start the project. While this may seem to be a very small work order, and it is, it was the largest that Mr. Yob could sign without further approvals. The politics of the situation was that the TRANSIT program, another satellite navigation system, was administered in the same office and Yob knew that the opposition to another navigation system would be fierce, and it was, but it became apparent only after this project was well on its way.

The Invention

What is the GPS invention? Recently awards have been made for “the invention of GPS”. It might be well to look at what makes GPS unique. I believe it is the means of measuring the range between the user and the various satellites. There are at least two techniques for measuring this range. The first, and most common, is to have a user interrogate the satellite, receive a return signal and use the delay between the interrogation and the return as a measure of the range. The problem with this technique is that it is easily saturated by a large number of users. When we consider the millions of users today, it would be completely useless.

The passive user technique uses a clock in each of the satellites. The clock is free running and adjusted as it varies from the standard time. The clock used in the initial passive satellites was controlled by a quartz crystal, as atomic oscillators, at the time, were too large for use in reasonably sized satellites. The problem with the quartz oscillators was that they were of marginal accuracy and that they could be affected by space particles. Never-the-less crystal clocks were used in the first two TIMATION launches and valuable information was obtained from them. When a small rubidium controlled oscillator was developed by Jeychart in Germany we immediately purchased a number of them and modified them for inclusion in the third TIMATION launch.

We had an interesting experience with the crystal oscillators. A good friend and accomplished scientist suggested that, instead of using an atomic oscillator, we use a large number of crystal oscillators and query them to get a best value. With 100 oscillators we should get a stability improvement of 10 or so. While the math was correct the problem with the crystals is that all of them can be zapped simultaneously in space.

The first TIMATION satellite was launched piggyback on May 31, 1967. It was reported in the AVIATION week & Space Technology issue of November 27, 1967. The article, by Philip J. Klass was titled “New satellite Navaid Tested Successfully.” In paragraph 3 we quote, “The time-navigation concept is based on a passive-ranging technique that does not require any transmissions by the aircraft, an important consideration for military use.” And later, “the concept requires that the satellite have an extremely accurate crystal oscillator, whose precision approaches that of an atomic standard.”

On October 6, 1969, p. 20 a description of the Air Force’s 621B system was contained in the same magazine. It was described further on February 2, 1970. An article titled “Satellite Clusters Studied for Tri-Service Navigation” was written by Barry Miller. I’ll quote from the forth paragraph: ”…or network of four or five 1,000-lb satellites to be orbited over the North American continent for 24-hour service to Western Hemisphere users.”

Just after Mr. Miller’s article, the Navy and Air force proposals were described in EASCON’ 69, a conference held at the Sheraton Park Hotel in Washington, DC on October 27-29, 1969. The conference had papers on a wide view of electronics but we can limit our interest to the session on SATELLITE NAVIGATION SYSTEMS. Papers in this session include one on Low Altitude Satellites by R. B. Kershner of the Johns Hopkins Applied Physics Laboratory, one on Medium Altitude Satellites by Roger Easton of the Naval Research Laboratory and two on 24-Hour Orbits by personnel from the Aerospace Corporation. The article by Dr. Kershner had little long-range impact so it will not be discussed further. The articles on MID-ALTITUDE SATELLITES had a huge impact and will be seen to form the basis of what became GPS.


By Roger L. Easton, U.S. Naval Research Laboratory

After looking at higher and lower altitudes the mid-altitude (approximately 1 earth’s diameter) polar circular satellite constellation has been selected as a prime possibility for an accurate, all weather, always available, three dimension. U.S. based navigation system. This paper also shows results from the first passive ranging satellite, the first TIMATION


This paper starts with a set of system properties that are desired for a high accuracy navigation system - all the properties are good and many would be met by the satellite system to be described were it not for fatal flaws in the constellation.

The paper describes the two methods of measuring range - the usual transponder-beacon type and the time-ordered satellite clocks (without attribution to the TIMATION original design).

The description of the satellite constellation is described as follows: “Synchronous orbits (not necessarily circular or equatorial) are preferred from the standpoint of providing good coverage and the ability to deploy less than a full global system. A particularly attractive configuration employs one satellite in a synchronous, near circular, equatorial orbit in conjunction with three or four satellites in inclined, elliptical, carefully phased orbits, having the property that their ground traces follow a common near-circular path around the ground trace of the first satellite. Such a constellation provides continuous regional geographic coverage with a nearly ideal geometry for providing accurate navigation. Three or four such constellations can provide nearly global coverage.”


The fatal flaws in this system are both connected to the elliptical orbits. The bigger flaw is that with certain elliptical orbits the satellites are unstable. If one remembers the Russian Molniya satellites he is struck by how short a time they stayed in orbit. The orbit was unstable and the satellite soon came to its end.

The second flaw with the elliptical orbits is that, if provided with stable clocks, the satellites would result in a much more complicated system. The obvious question that results is the following: Why not use circular orbits and not have the problems? Besides, circular orbits with high inclinations provide true global coverage.


While the optimum altitudes was discussed in the EASCON’69 paper a more analytical paper appeared in the May, 1970 issue on Naval Research Reviews. In this issue the use of the intercept method of Marcq St-Hilaire was discussed and examples were given. The advantage of a Line-of-Position type of navigation is that the calculations can be pre-computed and the results obtained almost instantly. It should be remembered that at that time the integrated circuits were not developed to the point where all the computations could be done in an inexpensive, handheld device.

I repeat the third paragraph; “The ability of USAF and Navy to resolve their long-standing differences over the orbital configuration by adopting basically the Navy-proposed constellation arrangement has eliminated one of the major obstacles to Pentagon approval for the program.”


The “Labor Day Conference” of 1973 was held in a motel on Spring Hill Road in Virginia (not in the Pentagon as various peoples have recalled). At this meeting Air Force Colonel Bradford Parkinson announced that the USAF-Aerospace proposal for System 621B was too expensive to complete. At the suggestion of the late Navy Captain (ret.) David Holmes, the Air force agreed to use and manage the technology of the NRL proposal. This idea was accepted and that is how GPS began.

Many people, including some who were directly involved and should know better, have reported a different genesis for GPS. Some of the misrepresentations were made in PHYSICS TODAY. Daniel Kleppner authored an article that was published in the January, 1994 issue (p. 9) where he states that “GPS was initiated by the Navy in the 1970s and taken over by the Air force in the 1980s” (his dates were incorrect by a decade, the Navy initiated the program in the 1960s and the Joint Program Office took over in the 1970s).

In response to Mr. Kleppner’s column, the late Dr. Ivan Getting wrote an article in the October 1994 issue entitled, “GLOBAL POSITIONING SYSTEM’S HIIDDEN HISTORY.” In this article, Dr. Getting evidently confused the GPS system with the earlier 621B proposal. mistakenly credited the cesium clocks to the Air Force (though they were developed for NRL and flown in a Navy satellite). I subsequently answered Dr. Getting’s letter in the December, 1995 issue that also included Dr. Getting’s response where he acknowledged that he had confused “GPS” with “621B” and that he meant “621B” when he spoke of “GPS.” Unfortunately, correction fatigue had set in by then and I let several inaccuracies in Dr. Getting’s article stand uncorrected.

Perhaps now is the time to revisit Dr. Getting’s letter in PHSICS TODAY. In the December, 1995 issue on page 92, appears this statement, “ After all, a third of a century has elapsed since the start in 1963 of the Aerospace Corporation efforts on what is currently called GPS and the system studies and demonstrations of the 1960s and GPS are basically the same.”

As we have already seen, the Aerospace effort, usually called 621B, differed in fundamental ways from the Navy effort called TIMATION. For starters, 621B would not have worked because many of its orbits were unstable. Second, 621B was not suitable for passive ranging because the frequencies of its oscillators would have varied as they went about their elliptical orbits. If used on an interrogation basis, its user population would be limited. The TIMATION proposal used circular orbits and a non-varying frequency due to gravitation. It would have worked because GPS does work and GPS is based on TIMATION. So it is a wild stretch of imagination to call 621B and TIMATION “basically the same”. As the courts might say, “The statement is ‘without merit.’”


The magazine “Aviation Week & Space Technology” ran several articles and boxes on navigation satellites through the years. Here are some of the references:

November 27, 1967 pp 63-64 “New Satellite Navaid Tested Successfully

The satellite was the first TIMATION

April 8, 1968 p. 25 “Hughes, TRW Teams to Study Navsat”

The Navsat to be studied was 621B.

March 31, 1969 p23 “Tactical Navsat Studies to Continue” again 621B

October 6, 1969 p20 “USAF Pushing Navigation Satellite System Study” again 621

February 2, 1970 pp 20-21 “Satellite Clusters Studied for Tri-Service Navigation” again 621Bv

May 8, 1972 pp 50-57 “USAF Studies Navsat Proposal” ¾ 621B with eggbeater patterns

August 20, 1973 page 65 “Plans for Defense Navsat Readied” ¾ 621B with rotating “Y”

November 26, 1973 pp 46-51 “Compromise Reached on Navsat”, and
“Resolve their long-standing differences over the orbital configuration by adopting basically the Navy-proposed constellation arrangement has eliminated one of the major obstacles to Pentagon approval for the Program.”

April 15, 1974, pp 22-23 “Satellite Navigation Network Defined”

The September, 1974 issue of Popular Science ran an article by Werner von Braun titled,
“24 satellites to give global pinpoint navigation”

Peter Galison’s book, “Einstein’s clocks, Poincare’s Maps” describes the GPS on pages 287-288. He also describes the use of cesium clocks to measure the effect of gravitation on the frequency of the clocks.

The EASCON ’69 conference was reported in 69 C 31-AES

Two patents are of interest. The first is by Roy Anderson of GE. In this patent he describes a proposed navigation system for NASA. He describes a satellite constellation much like the GPS array and years sooner. The patent number is 3,384,891 of May, 1968 and is titled “METHOD AND SYSTEM FOR LONG DISTANCE NAVIGATION AND COMMUNICATIONS,” filed February 11, 1965

The second patent of interest is titled, “NAVIGATION SYSTEM USING SATELLITES AND PASSIVE RANGING TECHNIQUES”, NUMBER 3,789,409 of Jan. 29, 1974 assigned to Roger L. Easton and filed October 8, 1970.